Yanyu Su

Online spatio-temporal Gaussian process experts with application to tactile classification

In this work, we are primarily concerned with robotic systems that learn online and continuously from multi-variate data-streams. Our first contribution is a new recursive kernel, which we have integrated into a sparse Gaussian Process to yield the Spatio-Temporal Online Recursive Kernel Gaussian Process (STORK-GP). This algorithm iteratively learns from time-series, providing both predictions and uncertainty estimates. Experiments on benchmarks demonstrate that our method achieves high accuracies relative to state-of-the-art methods. Second, we contribute an online tactile classifier which uses an array of STORK-GP experts. In contrast to existing work, our classifier is capable of learning new objects as they are presented, improving itself over time. We show that our approach yields results comparable to highly-optimised offline classification methods. Moreover, we conducted experiments with human subjects in a similar online setting with true-label feedback and present the insights gained.

A syntactic approach to robot imitation learning using probabilistic activity grammars

This paper describes a syntactic approach to imitation learning that captures important task structures in the form of probabilistic activity grammars from a reasonably small number of samples under noisy conditions. We show that these learned grammars can be recursively applied to help recognize unforeseen, more complicated tasks that share underlying structures. The grammars enforce an observation to be consistent with the previously observed behaviors which can correct unexpected, out-of-context actions due to errors of the observer and/or demonstrator. To achieve this goal, our method (1) actively searches for frequently occurring action symbols that are subsets of input samples to uncover the hierarchical structure of the demonstration, and (2) considers the uncertainties of input symbols due to imperfect low-level detectors. We evaluate the proposed method using both synthetic data and two sets of real-world humanoid robot experiments. In our Towers of Hanoi experiment, the robot learns the important constraints of the puzzle after observing demonstrators solving it. In our Dance Imitation experiment, the robot learns 3 types of dances from human demonstrations. The results suggest that under reasonable amount of noise, our method is capable of capturing the reusable task structures and generalizing them to cope with recursions.

Development of Search-and-rescue Robots for Underground Coal Mine Applications

Rescue missions after coal mine accidents are highly risky and sometimes impossible for rescuers to perform. To decrease the risk to rescuers, two generations of tracked mobile robots have been designed and developed to replace the rescuers. In this paper, we present the design iterations with experiments carried out in training sites for rescuers and in working coal mines, and we summarize the design and development experiences of the mobile robots for such rescue missions. In coal mine rescue robots, the explosion-proof and waterproof designs are adapted to the explosive and wet environments, while the suspension systems are adapted to the unstructured working environments. The design and development experiences may provide a reference for designing and developing future mobile robot systems for coal mine accident rescue missions.

Robust grasping for an under-actuated anthropomorphic hand under object position uncertainty

This paper presents a grasp execution strategy for grasping an object with one trial when there is uncertainty in the object position. This strategy is based on three grasping components: 1) robust grasp trajectory planning which can cope with reasonable amount of initial object position error, 2) sensor-based grasp adaptation, and 3) compliant characteristics of the under actuated mechanism. This strategy is implemented and tested on the iCub humanoid robot. Two experiments and a demo of the iCub robot playing the Towers of Hanoi game are carried out to verify our system. The results demonstrate that the iCub using this approach can successfully grasp objects under certain position error with its under-actuated anthropomorphic hand.

A morphable template framework for robot learning by demonstration: Integrating one-shot and incremental learning approaches

Robot learning by demonstration is key to bringing robots into daily social environments to interact with and learn from human and other agents. However, teaching a robot to acquire new knowledge is a tedious and repetitive process and often restrictive to a specific setup of the environment. We propose a template-based learning framework for robot learning by demonstration to address both generalisation and adaptability. This novel framework is based upon a one-shot learning model integrated with spectral clustering and an online learning model to learn and adapt actions in similar scenarios. A set of statistical experiments is used to benchmark the framework components and shows that this approach requires no extensive training for generalisation and can adapt to environmental changes flexibly. Two real-world applications of an iCub humanoid robot playing the tic-tac-toe game and soldering a circuit board are used to demonstrate the relative merits of the framework.

An under-actuated manipulation controller based on Workspace Analysis and Gaussian Processes

The kinematic modelling has been applied to many controllers of under-actuated manipulators. Most of these studies assume that the control process is conducted within the workspace. However, as such a kinematic model cannot describe the situations when the stable grasping is violated in the real environment, these controllers may fail unexpectedly. In this paper, we propose a combination of kinematics based Workspace Analysis (WA) and Gaussian Process Classification (GPC) to model the success rates of control actions in the theoretical workspace. We also use the Gaussian Process Regression (GPR) to model the residual between the prediction of the WA and the ground truth data. We then apply this integrated model, Gaussian Processes enhanced Workspace Analysis (GP-WA), into an optimal controller. The optimal controller is implemented on a planar under-actuated gripper with two three-phalanx fingers. Two sets of simulation experiments are carried out to validate our method. The results demonstrate that the optimal manipulation controller based on GP-WA achieves high control accuracy for manipulating a wide range of objects.

Enhanced kinematic model for dexterous manipulation with an underactuated hand

Recent studies on underactuated manipulation usually describe the system with a Kinematic Model (KM), which is built by adding external constraints to the standard manipulation analysis method. However, such external constraints are easily violated in a real-world dexterous manipulation task which results in significant control errors. In this work, the Enhanced Kinematic Model (E-KM), an integrated model of the KM and the Sparse Online Gaussian Process (SOGP) is proposed. The E-KM can compensate the shortfalls of the KM by on-the-fly training the SOGP on the residual between the prediction of the KM and the ground truth data. Based on the E-KM, we further contribute an optimal controller for underactuated manipulations. This optimal E-KM controller is implemented and tested on the iCub, a humanoid robot with two anthropomorphic underactuated hands. Two sets of real-world experiments are carried out to verify our method. The results demonstrate that the controller using E-KM statistically can achieve higher control accuracy than using solely using the KM for a wide range of objects.

A ROS/Gazebo based method in developing virtual training scene for upper limb rehabilitation

Virtual Reality (VR) presents a promising future in the field of rehabilitation due to its advantages brought to the training process as is indicated in many articles and researches. In this paper, we describe a novel method in developing a virtual training environment for a 5 degrees of freedom (DOF) upper limb rehabilitation robot which has been designed to help provide assistance for patients who survive stroke but remain hemiplegic to complete rehabilitation exercise. The method involved utilizes ROS (Robot Operating System) and Gazebo (a multi-robot simulator) to set up an interesting virtual scene of daily life in a 3D world to facilitate the patients to move their affected arms with synchronous visual feedback and interact with the virtual training task. The control method of human model in the virtual world and the communication mechanism between the host machine which controls movement of the real robot and the master machine which runs the VR will be introduced. Finally, a virtual training environment containing reaching task set in a modern kitchen is presented.

U-Pendant: A universal teach pendant for serial robots based on ROS

Many corporations have produced various teach pendants. However, most of these products are exclusive in terms of communication protocols and robot programming languages. Therefore, programs and teach pendant are usually not exchangeable between robots. To tackle this problem, we present U-Pendant, an open source universal teach pendant for serial robots based on Robot Operating System (ROS), a meta-operating system with a number of robotic applications shared online, and Yet Another Robot Controller (YARC), a universal kinematic controller for serial robots. We design and implement essential functionality and interactivity of teach pendants on U-Pendant. Several demonstrations are made to show the universality and functional superiority of U-Pendant. The source code and hardware design have been made publicly accessible for the community.

Increasing the Accuracy and the Repeatability of Position Control for Micromanipulations Using Heteroscedastic Gaussian Processes

Many recent studies describe micromanipulation systems by using complex Analytic Forward Models (AFM), but such models are difficult to build and incapable of describing unmodelable factors, such as manufacturing defects. In this work, we propose the Enhanced Analytic Forward Model (EAFM), an integrated model of the AFM and the Heteroscedastic Gaussian Processes (HGP). The EAFM can compensate the shortfalls of the AFM by training the HGP on the residual of the AFM. This also allows the HGP to learn the repeatability of the micromanipulation system. Based on the EAFM, we further contribute an optimal position controller for improving the accuracy and the repeatability. This optimal EAFM controller is implemented and tested on a three degree-of-freedom micromanipulator based micromanipulation system. Two sets of real-world experiments are carried out to verify our method. The results demonstrate that the controller using EAFM can statistically achieve higher accuracy and repeatability than solely using the AFM.